Apologies if this is a duplicate of another post - I couldn't find an answer on the forum or in the docs...
On the host machine (running proxmox), I've imported a zfs pool (let's call it 'TANK') that I moved from a previous build (FreeNAS) using the following command in proxmox:
zpool import -f TANK
It's made from 10 2.7TB disks in RAIDZ2 configuration. It contains about 7TB of media files that I want to be able to access from a guest running ubuntu server 20.04 LTS.
My question is, what's the 'best' way to do that? By 'best', I'm aiming for a solution that:
1) Doesn't lose any of the information already on TANK
2) Makes the existing files and directories on TANK available to the ubuntu host for it to create, read, update and delete
3) Doesn't compromise the data integrity delivered by zfs
4) Maximises read and write speeds
I've successfully added this pool as storage within the GUI...
'Datacenter'>'Storage'>'Add'>'ZFS', with options
'VM 100'>'Hardware'>'Add'>'Hard disk', with options
...
Disk /dev/vda: 19.54 TiB, 21474836480000 bytes, 41943040000 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
...
And the relevant /etc/fstab entry in the guest (ubuntu server OS) is:
'/dev/vda /TANK ext4 defaults 0 0'
When I run 'sudo mount -a', I get the error:
mount: /TANK: wrong fs type, bad option, bad superblock on /dev/vda, missing codepage or helper program, or other error.
In case it's relevant, I don't use a ZIL, L2ARC etc. The pool was created with 'zpool create TANK raidz2 /dev/sda /dev/sdb ... /dev/sdj'
Help please.
On the host machine (running proxmox), I've imported a zfs pool (let's call it 'TANK') that I moved from a previous build (FreeNAS) using the following command in proxmox:
zpool import -f TANK
It's made from 10 2.7TB disks in RAIDZ2 configuration. It contains about 7TB of media files that I want to be able to access from a guest running ubuntu server 20.04 LTS.
My question is, what's the 'best' way to do that? By 'best', I'm aiming for a solution that:
1) Doesn't lose any of the information already on TANK
2) Makes the existing files and directories on TANK available to the ubuntu host for it to create, read, update and delete
3) Doesn't compromise the data integrity delivered by zfs
4) Maximises read and write speeds
I've successfully added this pool as storage within the GUI...
'Datacenter'>'Storage'>'Add'>'ZFS', with options
- 'ID': TANK
- 'ZFS Pool': TANK
- 'Content': Disk image
- 'Nodes'=>All (No restrictions)
- 'Enable': selected (checked)
- 'Thin provision': selected (checked)
- 'Block Size': 4k
'VM 100'>'Hardware'>'Add'>'Hard disk', with options
- Bus/Device: VirtIO Block
- Storage: TANK:vm-100-disk-2
- Disk size (GiB): 20000
- Cache: Default (No cache)
- Discard: not selected (unchecked)
- SSD emulation: not selected (unchecked)
- IO thread: not selected (unchecked)
- Read limit (MB/s): unlimited
- Write limit (MB/s): unlimited
- Read limit (ops/s): unlimited
- Write limit (ops/s): unlimited
- Backup: not selected (unchecked)
- Skip replication: not selected (unchecked)
- Read max burst (MB): default
- Write max burst (MB): default
- Read max burst (ops): default
- Write max burst (ops): default
...
Disk /dev/vda: 19.54 TiB, 21474836480000 bytes, 41943040000 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
...
And the relevant /etc/fstab entry in the guest (ubuntu server OS) is:
'/dev/vda /TANK ext4 defaults 0 0'
When I run 'sudo mount -a', I get the error:
mount: /TANK: wrong fs type, bad option, bad superblock on /dev/vda, missing codepage or helper program, or other error.
In case it's relevant, I don't use a ZIL, L2ARC etc. The pool was created with 'zpool create TANK raidz2 /dev/sda /dev/sdb ... /dev/sdj'
Help please.
Last edited: